4 ms·

You said "And you can't do it with solar." He says it's possible in the very video you pointed at, his exact words "... unless we have subsequent power generat

by guerby 5y ago

You said "And you can't do it with solar." He says it's possible in the very video you pointed at, his exact words "... unless we have subsequent power generation at houses". EV are the perfect demand shifting and averaging tool for the grid, which is why Elon Musk says what he says, you can reduce peak demand on the grid this way. All that is in the report I cited, multi year study by lots of engineers on the case of the French grid. Now of course the US electrical grid is not in real good shape compared to other developped countries, but full EV and end of fossil fuels is at least 3 decades away. Australia went from zero to 25% of homes with solar in a decade, 2.4 times increase in the last five year according to https://en.wikipedia.org/wiki/Solar_power_in_Australia https://en.wikipedia.org/wiki/Solar_power_in_Australia For yearly energy: USA vehicules drives about 3200000 millions miles per year, at 0.25 kWh/mile (TM3 LR) that's 800 billions kWh. Current electricity production in USA is 4000 billions kWh. Plus twenty percent is not that much.

robomartin 5y ago
Well, he sells solar panels and batteries. That's all I'll say. It isn't that it is impossible with solar. The problem is that the massive scale of the necessary deployment in both panels and batteries will be far, very far, from "green". The scale of manufacturing and mining we would have to engage in would make any environmentalist vomit. > USA vehicules drives about 3200000 millions miles per year, at 0.25 kWh/mile (TM3 LR) that's 800 billions kWh. Current electricity production in USA is 4000 billions kWh. Plus twenty percent is not that much. This is wrong. Why would Elon (and myself, starting many years ago) say we need to build an additional 900 to 1400 GW of generation capacity if all we need is 20% more? One of the problems with simplified/generalized calculations is what I call the "physics cow problem". When you study physics we tend to simplify problems and eliminate variables (no friction, no air, no gravity, etc.) in order to be able to study the fundamentals. That's where, all of a sudden, a cow goes from being a highly complex organism to "assume a cow is a uniform sphere of milk one meter in diameter". Well, you can't do that when you are talking about 300 million vehicles. You have to model time zones, work hours, miles driven, charging power/energy requirements, multimodal transportation (motorcycles, small cars, large cars, trucks, large trucks, massive trucks, forklifts, trains and whatever else you can imagine). You then have to create random-but-reasonable utilization profiles. Make enough of them to try to simulate the thousands of use-cases for vehicles. From the parent taking kids to school every day to the Uber driver, soccer mom, travelling sales person, warehouse operation, etc. To this you have to add charging behaviors. People and companies are not going to charge their vehicles exactly the same way and at the same time. Some will want (or need) to fast-charge. Others will have the ability to slow charge during the day. And yet others will fast or slow charge at night. Some will have a routine, while others might appear to be more random due to utilization profile. Anyhow, once you create enough variables and variability you can start to attach numbers and run simulations with various fleet characteristics over time and geography. Not easy, yet this is the only way to truly get a sense of proportion. You can't say "a cow is a uniform sphere of milk one meter in diameter" and get answers you can stake your reputation on. As for the grid. Well, it needs a major upgrade. I can't speak for France or any other nation other than to say that I would not be surprised to learn that most countries have the same problem: The grid was designed to deliver power to cities, towns, homes, businesses and schools. Given that most cities of the world don't expand or change rapidly, it is likely that most power distribution systems were implemented with a little excess capacity, but not much. For example: No politician would have survived proposing that my neighborhood, which is about 25 years old, be built with twice the required power delivery capacity. Older parts of the world likely have it worse. A transition to electrics will have to use electric energy to provide a portion of the energy now delivered through fossil fuels. That is a major step-change in energy delivery to homes and businesses. Maybe China is setup to handle this. I sincerely doubt most nations of the world are anywhere close to being able to manage it.

by guerby 5y ago

robomartin 5y ago

Well, he sells solar panels and batteries. That's all I'll say. It isn't that it is impossible with solar. The problem is that the massive scale of the necessary deployment in both panels and batteries will be far, very far, from "green". The scale of manufacturing and mining we would have to engage in would make any environmentalist vomit. > USA vehicules drives about 3200000 millions miles per year, at 0.25 kWh/mile (TM3 LR) that's 800 billions kWh. Current electricity production in USA is 4000 billions kWh. Plus twenty percent is not that much. This is wrong. Why would Elon (and myself, starting many years ago) say we need to build an additional 900 to 1400 GW of generation capacity if all we need is 20% more? One of the problems with simplified/generalized calculations is what I call the "physics cow problem". When you study physics we tend to simplify problems and eliminate variables (no friction, no air, no gravity, etc.) in order to be able to study the fundamentals. That's where, all of a sudden, a cow goes from being a highly complex organism to "assume a cow is a uniform sphere of milk one meter in diameter". Well, you can't do that when you are talking about 300 million vehicles. You have to model time zones, work hours, miles driven, charging power/energy requirements, multimodal transportation (motorcycles, small cars, large cars, trucks, large trucks, massive trucks, forklifts, trains and whatever else you can imagine). You then have to create random-but-reasonable utilization profiles. Make enough of them to try to simulate the thousands of use-cases for vehicles. From the parent taking kids to school every day to the Uber driver, soccer mom, travelling sales person, warehouse operation, etc. To this you have to add charging behaviors. People and companies are not going to charge their vehicles exactly the same way and at the same time. Some will want (or need) to fast-charge. Others will have the ability to slow charge during the day. And yet others will fast or slow charge at night. Some will have a routine, while others might appear to be more random due to utilization profile. Anyhow, once you create enough variables and variability you can start to attach numbers and run simulations with various fleet characteristics over time and geography. Not easy, yet this is the only way to truly get a sense of proportion. You can't say "a cow is a uniform sphere of milk one meter in diameter" and get answers you can stake your reputation on. As for the grid. Well, it needs a major upgrade. I can't speak for France or any other nation other than to say that I would not be surprised to learn that most countries have the same problem: The grid was designed to deliver power to cities, towns, homes, businesses and schools. Given that most cities of the world don't expand or change rapidly, it is likely that most power distribution systems were implemented with a little excess capacity, but not much. For example: No politician would have survived proposing that my neighborhood, which is about 25 years old, be built with twice the required power delivery capacity. Older parts of the world likely have it worse. A transition to electrics will have to use electric energy to provide a portion of the energy now delivered through fossil fuels. That is a major step-change in energy delivery to homes and businesses. Maybe China is setup to handle this. I sincerely doubt most nations of the world are anywhere close to being able to manage it.